Abstract: Against the backdrop of promoting the inheritance, innovation, and development of traditional Chinese medicine (TCM) and the establishment of the "three-combination" review evidence system integrating "TCM theory, human use experience, and clinical trials", hospital-prepared TCM formulations have become a vital source for transforming new TCM drugs. Standardized collection of clinical practice data from human use experience of these formulations plays a key role in summarizing such experience and facilitating research and development transformation. Based on policy requirements and methodological guidelines, this paper identifies the core elements of data collection, covering key dimensions such as target population characteristics, study factors, outcome variables, and covariates. It also proposes the use of causal directed acyclic graphs (DAG) to guide variable identification and collection. Additionally, this study emphasizes key methodological requirements including a focus on regulatory compliance in information gathering and structured processes in data acquisition, ensuring the standardization of TCM diagnostic information collection, and promoting the routine documentation of out-of-hospital outcomes and competing endpoints, along with the dimensional integration of the biopsychosocial medical model. This provides methodological support for constructing high-quality human use experience datasets, thereby facilitating evidence-based transformation of hospital-prepared TCM formulations and the development of new TCM drugs.